← registry

Evidence profile / source review

Aaru

Aaru describes a multi-agent simulation and behavioural modelling platform that generates populations for prediction and decision support.

Classification

Core buyer overlap

Population simulationBehavioural modelling

Publicly described methods

  • Multi-agent simulation
  • Population modelling
  • Prediction

Disclosure and limitation

The public source documents positioning and described methods. It does not by itself establish predictive validity or fitness for a specific decision.

Sources

  1. [1]
    Aaru vendor evaluation summary

    Aaru · first-party · accessed 2026-10-15

  2. [2]
    Aaru product site

    Aaru · first-party · accessed 2026-10-15

Evidence dossier / current interpretation

Claim-source review /

A specific first-party human-comparison or validation claim was located in the reviewed sources and recorded separately. This source-claim review does not establish human-reference validation or independent reproduction.

This review covers the cited sources, not every possible claim or unpublished study. “No specific claim located” does not mean no claim exists.

What this profile establishes

The cited sources establish how the organization publicly describes the tool and its methods.

What it does not establish

Unless explicitly established above, this profile does not establish fidelity to human populations, behavioural prediction, independent reproducibility, or suitability for a decision.

Vendor-reported: human-reference validation

Aaru reports comparisons against published survey responses across 2,993 questions and 186 studies, including observed mean TVD of 7.62 points and sample-adjusted MAE of 2.78 percentage points. These are vendor-reported evaluation results, not an SRI-established validation or independent reproduction.

Inspect the vendor source

This check is not established by SRI. A first-party report is not independent reproduction.

Correction right

See a factual error or stronger evidence?

Submit the exact claim, a public source, and your relationship to the tool. Corrections are logged; editorial conclusions remain independent.

Submit evidence